Edaphic adaptation

Some of the most remarkable adaptations in plants include adaptation to extreme edaphic stress. Therefore we are applying genome scanning approaches to understand the solutions evolution offers in some of the many independently-evolved lineages of metal, drought, and serpentine-tolerant plants. Our initial results point to broadly orchestrated, polygenic responses to these selective pressures. Concerted changes indicate that even in the case of an environmetnal pressure, well-orchestrated internal adaptations may be marshalled, calling for a better understanding of internal adaptation (i.e., systems-based analysis of compensatory changes in the genome). In addition, these first results are pointing to striking instances of repeated evolution and gene flow between species that may mediate some of these adaptations.


Why Serpentine? 

Serpentine soils present a multidimensional hazard to plant life. Not only do they offer marginal levels of essential nutrients such as Ca, P, K, and N, but they are also usually very porous, with a high propensity toward drought. Low Ca:Mg ratios are a defining feature of serpentine environments and result in very low Ca uptake. These insults are typically compounded by the presence of phytotoxic levels of heavy metals such as Ni, Cd, and Zn, which leads to stunting and chlorosis, along with antagonistic effects on Fe uptake. As a result, serpentine environments are characterized by minimal ecosystem productivity and high rates of endemism. Evolution has nevertheless forged populations that thrive among these stresses, which by adapting to this ‘serpentine syndrome,’ suggest solutions to challenges in crop improvement.

In contrast to our recent work on adaptation to genome duplication in the same A. arenosa system, we discovered a relatively diverse array of genes implicated in serpentine adaptation, from strong sweeps in dehydration tolerance coding loci (ERD4 below), to loci involved in sulfur transport (SULTR1;1), metal transport, and root growth. We see clear selective sweeps in many categories consistent with adaptation to ‘serpentine syndrome’: dehydration tolerance, ion transport (Ca, Mg, and K transport-related genes), stress, and root branching and growth.

Example sweeps in nutrient uptake and drought stress candidates in serpentine-adapted A. arenosa

(A) Allele frequency differences (AFD) in two example differentiated regions. Dots represent polymorphic SNPs. X-axis gives chromosome location. Y-axis gives degree of differentiation calculated by plotting the absolute value of the difference in AFD between serpentine-adapted and non-serpentine comparison groups. Arrows indicate gene models. Red arrow indicates sweep candidate with localized differentiation. These candidates, EARLY RESPONSE TO DEHYDRATION-LIKE 4 (ERD4) and SULPHATE TRANSPORTER 1;1 (SULTR1;1), have predicted gene functions are concordant with the challenges surrounding serpentine adaptation.

(B) Several population differentiation metics plotted in the ERD4 region. X-axis gives chromosome position. Red axis = Fst. Yellow axis = 2dSFS. Blue axis = Diversity/Differentiation residual metric. Colored dotted lines give 0.5% cutoff for each respective metric. Combined with the AFD plots in (A), these indicate that the ERD4 gene coding locus has undergone a selective sweep specifically in the serpentine-adapted population. We see 58 gene coding loci scattered throughout the genome that are all extreme outliers for all of the metrics shown here, indicating that many genes have coevolved together to allow these plants to endure the multiple hazards of colonizing serpentine environments.

In recent genome scans we see clear signatures of selection, typically consisting of single gene peaks of high differentiation, falling off immediately following the coding locus, due to negligible linkage in A. arenosa. This makes identification of candidate genes unambiguous. In contrast to most (if not all) other genome scans, most of our top selective sweep candidates have serpentine-relevant adaptive functions (from published studies, not simply GO assignments): e.g., dehydration tolerance genes, heavy metal transporters, and root macronutrient uptake transporters.

We interpret these results as showing us not only adaptations to external challenges, but also compensatory adjustments to the changed physiological state of the cell (as in Yant et al 2013, where we observed coevolution of interacting meiosis proteins). That these serpentine-tolerant lineages are polyploid is an important hint toward possible mechanisms of preadaptation.


The work above is part of a set of ongoing collaborations with David Salt (Aberdeen, Scotland, U.K.).

We also have a set of exciting toxic mine adaptation projects in collaborations with Ute Krämer (Bochum, Germany)